Premature mortality among males with substance use disorders

Previous studies have demonstrated that substance use disorders (SUD) are associated with premature mortality in clinical populations, though findings regarding the effect of antisocial personality disorder (ASP) on mortality among persons with SUD are less clear. However, it is unclear to what extent that finding generalizes to persons with SUD in the community, because very little work has been done involving mortality in longitudinal studies of mortality associated with SUD. The objective of this study was to assess whether the presence of SUD or ASP is associated with early mortality among males (fathers) in a predominantly community sample, using a 15-year prospective longitudinal study design.

We conducted a prospective longitudinal study of adolescents and their fathers. The adolescent subjects were recruited at age 10-12 years, with follow-up evaluations at ages 14, 16, 19, 22, and 25. Questions were asked about paternal mortality during each of those visits. The study sample for this study was the 769 fathers of the adolescent subjects, who included N=341 fathers with a DSM-III-R diagnosis of SUD and N=428 control fathers without a SUD. 89% of these fathers were recruited from the community, and 11% were recruited from clinical sources. Comorbidity patterns were described. A multivariate Cox regression analysis was performed with the father's age at death or last assessment as the dependent variable, and education, SUD, and ASP as the independent variables.

Lower education level, the presence of a substance use disorder, and the presence of antisocial personality disorder were significantly associated with earlier mortality (Wald=5.1, df=1, p=0.024; Wald=5.1, df=1, p=0.024, and Wald=5.5, df=1, p=0.019, respectively). Most subjects died from medical illnesses, as opposed to drug overdoses or accidents, which is different from the pattern often noted in clinical samples.

The results of this study demonstrate that the presence of SUD, the presence of ASP, and a lower education level were associated with early mortality in our primarily community-based sample, which extends previous reports of similar findings in clinical samples. The magnitude of the prematurity of the deaths was less that that generally noted in previous studies involving clinical samples, and the causes of death were also somewhat different from those noted in clinical samples. The majority of cases of mortality in our SUD sample resulted from medical illnesses rather than from accidents or overdoses.